The Impact of Influenza

Estimates of the impact of influenza on the U.S. population indicate that the disease has major significance. Influenza surveillance in Houston, Texas, during the 1970s and 1980s recorded an influenza infection rate of 33 per 100 among families with young children; approximately two thirds of infected persons exhibited illness (1). Rates of infection among children and adolescents in Tecumseh, Michigan, during this time were similar (4). Although the estimate of medically attended cases is crude, combining the Houston experience with national data suggests that an average of 25 million persons with influenza seek health care in the United States each year (1, 7). The validity of these estimates is supported by recent monitoring of two health maintenance organizations in central Texas. During the influenza epidemic period, rates of visits related to acute respiratory illness were similar to rates seen from 1974 to 1985 (Glezen WP. Personal communication). The Centers for Disease Control and Prevention estimate that between 114 000 and 146 000 persons are hospitalized each year because of influenza. Assuming that Houston is representative of the entire country, virologic surveillance of persons hospitalized with any acute respiratory illness from 1991 to 1995 provides a national estimate of 257 500 hospitalizations attributable to influenza (3, 5, 8). Estimates of associated deaths have varied between 20 000 and 40 000 per year (1, 2, 5).

Although exact tabulations of illnesses and complications attributable to influenza virus infection are not available, the preceding estimates indicate that the morbidity and mortality caused by influenza are major health problems. It is well known that school-age children and young adults have the highest rates of influenza infection and illness (1, 3, 4). Hospitalization rates are highest among infants and older persons, while mortality rates are highest among elderly persons (1-3, 5, 6).

Current Ways To Control Influenza

Vaccines

Inactivated Vaccine

The current available vaccine is inactivated and contains two subtypes of the influenza A virus (H1N1 and H3N2) and an influenza B virus (2). The vaccine is recommended for those at risk for severe influenza, hospitalization, and death and for those most likely to transmit virus to such persons (Table 1) (2). Recently, vaccination was also recommended for pregnant women and persons between 50 and 64 years of age. Although pregnancy was recognized as a risk factor during influenza pandemics, pregnant women were added to the recommended vaccine recipients because of data indicating that they were at risk during the annual epidemic (9). The age recommendation was reduced from 64 years to 50 years because persons 50 to 64 years of age often have underlying disease, and vaccination rates in this group have been low (2). Vaccination, which is primarily intended to prevent the complications of influenza, is now recommended for almost half of the U.S. population.

The influenza vaccine has low reactogenicity and effectively prevents influenza-specific illness in 70% to 90% of healthy adults when the vaccine and epidemic viruses match antigenically (2, 10, 11). Because antigenic variation with time is characteristic of the influenza viruses, protection frequencies are lower when the vaccine and epidemic viruses do not match; in addition, they are generally lower among very young and elderly persons, whose immune responses tend to be weaker (2, 12, 13). Protection against pneumonia, hospitalization, and death, however, remains high among elderly persons (2, 12, 13). Overall, during epidemic periods, both death from any cause and hospitalization related to respiratory disease and congestive heart failure have generally decreased by 30% to 50% among vaccinated older persons (14). Respiratory illness, physician visits, and sick leave among vaccinated healthy adults have also decreased by 30% to 50% (15).

Despite extensive evidence that the influenza vaccine effectively prevents disabling and serious disease among the high-risk persons for whom it is recommended, implementation of the recommendations has only recently improved (2). In 1998, vaccination of elderly persons (persons 65 years of age) exceeded 70% for the first time, and the goal for 2010 has been set at 90%. Vaccination of high-risk persons younger than 65 years of age has lagged: In 1997, fewer than 30% of younger persons with heart disease, lung disease, or other chronic disorders were vaccinated. Similarly, vaccination of health care personnel who could transmit influenza to high-risk persons is well below desirable levels. Renewed efforts to improve immunization numbers are needed, particularly among practicing physicians, who deliver most of the vaccine.

Attenuated Vaccine

A live attenuated vaccine that is administered by nasal spray is nearing availability. It has been shown to be safe, immunogenic, and efficacious in several clinical trials (1, 2, 16, 17) and was approximately as efficacious as inactivated vaccine in studies among young adults (1, 17). In a multicenter study in children 1 to 5 years of age, it provided up to approximately 90% protection against influenza-specific illness during the subsequent winter. Persons who received the vaccine also had 30% fewer cases of otitis media and 21% fewer febrile illnesses (16). Revaccination the subsequent winter protected 86% of children from an antigenically "drifted" virus. In these studies, 10% to 20% of vaccine recipients had minor upper respiratory symptoms that were attributable to the vaccine (16, 17). Availability and use of this vaccine should substantially reduce the effects of influenza in children. Since children are thought to have a primary role in spreading infection, the overall magnitude of the epidemic could also decrease (1, 3).

Antiviral Agents

Amantadine and Rimantadine

Amantadine and rimantadine are related antiviral agents that have a similar mechanism of action and effectiveness for influenza A (Table 2). Each compound is effective for prevention and treatment of influenza A virus infections but ineffective for influenza B virus infections. Prophylaxis generally prevents approximately 50% of infections and reduces the intensity of many infections, thereby preventing 70% to 90% of illnesses (18). This level of protection is approximately the same as that conveyed by the inactivated vaccine. Both compounds are also effective for treatment of influenza A if treatment is begun within 48 hours of illness onset. In general, these agents shorten illnesses by approximately 1 day (19, 20).

Amantadine and rimantadine are given orally. Both drugs can cause nausea and vomiting, but the frequency of these adverse effects is low (21, 22). Amantadine stimulates catecholamine release, which presumably explains the occurrence of symptoms related to the central nervous system, such as anxiety, depression, and insomnia, in 10% to 15% of persons. Hallucinations and seizures can occur with high blood levels. Side effects are dose related for both drugs and resolve with discontinuation. Antiviral resistance develops rapidly during treatment with either drug. Persons who develop resistance do not exhibit a prolonged illness or an illness rebound but can transmit resistant virus to other persons (23). Illnesses are similar in persons infected with resistant virus and those infected with susceptible virus.

Zanamivir and Oseltamivir

Zanamivir and oseltamivir are related antiviral agents with a similar mechanism of action and effectiveness for both influenza A and B viruses (24-32) (Table 2). They were designed to inhibit the neuraminidase on the surface of the virus and can prevent as well as reduce the intensity of infection and the occurrence and severity of illness. Zanamivir is approved for treatment of influenza in persons at least 7 years of age who have been symptomatic for less than 2 days; oseltamivir is approved for treatment in those at least 18 years of age, but approval for use in those at least 1 year of age is pending (24, 25). In clinical trials, "clinical influenza" disappeared 1 to 1.5 days earlier in treated patients than in those who received placebo. Still greater benefit was seen in more severely ill patients and in those who started treatment on the first day of illness (26, 27, 30, 32). In some studies, frequency of such complications as otitis, sinusitis, purulent bronchitis, and antibiotic use was reduced.

Both zanamivir and oseltamivir have been shown to be effective for prevention if given before a community exposure (28, 29, 31, 32). In healthy persons, the two drugs have similar effectiveness for prevention of infection (30% to 50%) and illness (70% to 90%). A high degree of prevention among exposed family members (80% to 90%) has been seen with both drugs. In addition, a trial with oseltamivir among vaccinated residents of nursing homes showed a 92% reduction in influenza (31, 32).

Oseltamivir is given orally, and zanamivir is given through the mouth by inhalation vigorous enough to ensure a high flow rate. Zanamivir has been shown to be generally safe for use in asthmatic patients. However, because reduced flow rates have been seen in patients with chronic airway disease, caution is advised in these populations (24). Oseltamivir caused nausea and vomiting in a small percentage of patients (25). Approximately 1% of treated patients develop resistance to oseltamivir, but resistance to zanamivir has not yet been detected in healthy persons (31, 32).

Recommendations for Use

Recommendations for use of antiviral agents are given in Table 3. Antiviral agents are valuable alternatives to vaccine for prevention and for augmenting protection among high-risk groups, particularly patients who are immunocompromised and respond poorly to vaccine. They are also useful for preventing influenza in families and in institutional settings where influenza is known to spread easily (2). All high-risk persons with influenza and otherwise healthy persons with severe influenza should be treated because treatment shortens the duration of illness and, according to available data, reduces complication rates. All otherwise healthy persons with severe influenza should be treated, and treatment should be considered for all persons with clinical influenza. Treatment may also reduce disease spread, as has been seen in families in which only the index case received amantadine or rimantadine (1). Although treatment with an antiviral agent is not known to be effective, it is recommended for persons with influenzal pneumonia and for high-risk persons not seen until day 3 or 4 of illness.

At present, only amantadine and rimantadine are approved for prophylaxis, but zanamivir and oseltamivir should be approved for this purpose soon. The latter two compounds have the advantage of effectiveness against influenza B but the disadvantage of a higher cost. When prophylaxis against influenza A is desired, amantadine and rimantadine are recommended because they seem to be as effective as and are much less expensive than the neuraminidase inhibitors. Development of resistance in this circumstance has not been a problem. Rimantadine is preferred for safety, but amantadine is preferred if cost is an important consideration. Either zanamivir or oseltamivir must be used if prophylaxis against influenza B is desired. The two drugs are equally effective and are of similar cost, but ease of administration favors oseltamivir. Administration of zanamivir requires good physical coordination, lung function, and understanding of the inhalation maneuver.

Zanamivir and oseltamivir are preferred for treatment because they are effective against influenza B, and data suggest that these agents cause fewer complications and carry a lower risk for resistance. Since this risk increases with time and no data suggest that treatment lasting more than 5 days is beneficial, treatment should probably not exceed 10 to 14 days regardless of the patient's clinical status. When both treatment and prophylaxis are used during an influenza A epidemic in such population groups as families or patients in long-term care facilities, treatment with zanamivir or oseltamivir and prophylaxis with amantadine or rimantadine are recommended. The neuraminidase inhibitors have been shown to reduce complications and are not likely to induce resistance (31, 32). Moreover, if resistance develops and is transmitted to others, the virus would be sensitive to amantadine and rimantadine. Comparisons of the different antiviral agents in clinical trials could alter these recommendations.

Needs for Improved Control

Vaccines

The current inactivated vaccine for the influenza virus is minimally reactogenic and moderately immunogenic and offers substantial protection against influenza, particularly severe disease and complications (1, 2, 14, 15, 33). However, the protection is incomplete, and better vaccines are needed. The essential component of immunity to influenza is serum IgG antibody to the hemagglutinin surface protein (33, 34). This antibody class, which is predominant in lower respiratory tract secretions (the presumed sites for initiation of most infections) can prevent as well as reduce the intensity of infection. The IgA antibody can contribute to immunity, particularly in the upper respiratory tract, where it is the dominant immunoglobulin class (35). Other immune agents that can prevent or modify infection are antibody to the neuraminidase in serum and secretions; antibody to the matrix 2 protein, which can reduce the intensity of infection; and cytotoxic lymphocytes that primarily react with epitopes on the nucleoprotein or matrix protein and promote recovery from infection [36-39]. Therefore, on the basis on present knowledge, optimal vaccination can be said to have five goals: 1) simulation of serum IgG antibody to the hemagglutinin in adequate quantities in all vaccinated persons, 2) stimulation of IgA and IgG antibody to the hemagglutinin in upper respiratory secretions, 3) stimulation of antibodies of the same class in the same locations to the neuraminidase protein, 4) stimulation of an increase in the precursor cell population of cytotoxic lymphocytes, and possibly 5) simulation of antibodies to the matrix 2 protein. In addition to providing optimal immunity to homotypic viruses during the epidemic, this spectrum of responses should provide enhanced protection against viruses that are antigenically changed.

The immunologic needs for improving vaccines are known. The difficulty lies in developing new approaches and products that optimally induce the desired responses. Each of the responses is frequently stimulated to optimal levels by natural infection; however, such stimulation by the live attenuated vaccine is not consistently seen in older persons, presumably because the infection is of low intensity (40). A variety of approaches for improving immune responses and level of protection after vaccination have been described and are under development (33). With adequate effort, modern virologic and immunologic methods should permit development of vaccines that have the desired attributes. This will require a commitment by industry, government, and academia.

Antiviral Agents

When zanamivir and oseltamivir are approved for prophylaxis, options for application will be available for all circumstances and for all persons at least 1 year of age. Nevertheless, effectiveness is yet to be demonstrated for many groups and circumstances. Treatment must be effective for patients who are seen more than 2 days after onset, particularly those who will be hospitalized when first seen, those receiving intensive care, those with influenzal pneumonia, immunocompromised persons, elderly patients, and patients with underlying lung disease. The most immediate possibility for increasing effectiveness in these difficult circumstances is simultaneous use of zanamivir or oseltamivir and amantadine or rimantadine; these combinations have enhanced antiviral effects in vitro and in the mouse model (31). Despite options for increasing the value of current antiviral agents, a third class of anti-influenzal drugs would further enhance options and flexibility.

Rapid Diagnosis

For optimal application of antiviral prophylaxis, both community viral surveillance and the ability to diagnose rapidly are needed. Community surveillance allows the physician to know when influenza enters and leaves the community and whether it is influenza A or B or both. Rapid diagnosis permits a specific diagnosis in ill persons and can contribute to community viral surveillance. During the epidemic period, rapid diagnostic methods are not necessary in the clinic for initiating antiviral therapy of persons with typical febrile influenza. However, they can aid in making decisions, such as whether to withhold antibiotic treatment, for patients with other acute respiratory syndromes. A rapid test should be used for all patients hospitalized with an acute respiratory illness during the epidemic period. A positive test result provides both a diagnosis and data for infection control, guides therapy, and helps the clinician assess prognosis in a given case. Current tests have high sensitivity among children hospitalized with acute influenza but relatively low sensitivity in some populations, such as adults presenting to a clinic. Improved sensitivity among all groups of patients is needed (41).

The Challenge of Application

Control of influenza could immediately improve with optimal application of existing options and resources. Increasing current vaccination levels to more than 90% among all recommended groups should noticeably decrease the impact of influenza. Increased use of antiviral agents for treatment and short-term prophylaxis in exposed groups and increased longer-term use for community exposure of less responsive vaccinated persons would further reduce influenza's effects. To do this regularly will require coordinated action by practicing physicians and public health authorities, but the reward should be gratifying.

Reducing influenza to a minor medical problem will require approval of the live attenuated vaccine, development of better inactivated vaccines, availability of better rapid diagnostic methods, additions to the current antiviral options, and optimal application of all available methods. Vaccines need to be effective for more than 1 year, and antiviral agents and rapid diagnosis must be readily available at acceptable prices. Although decreasing the effects of influenza is potentially difficult, it is feasible given our current knowledge and capabilities. The way is there; what is needed is the will. Perhaps we will soon be able to master this uncontrolled disease of antiquity.